3800 stack member failure but no SNMP alert!

[DDGRUS]

On one of our 3800 switch stacks we recently had a member completely fail!

Luckily we had a spare so this was configured for the stack and connected up, very handy as the commander holds the full running config so you only do a config to get it to join the stack.

The issue we had was that the SNMP monitoring appeared to fail! We run 3 different monitoring tools

* SolarWinds

* HPE IMC

* Random software called SNMP manager

None of the following had picked up that a member in the stack failed! Just wondered if anyone has had anything similar happen, I am also having issues explaining the problem with HP! They keep trying to send me to IMC support

[Richard Litchfield]

It is always worth adding syslog to the monitoring tools, eg:

logging 10.20.30.185

this sends all entries that you see when doing a "show logging" to the syslog server. You would have seen some messages then! SNMP may or may not always capture that sort of failure. Some SNMP elements can be disabled from sending a trap (and be disabled by default - like loop-protect).